SKEW in Collabora Office: Calculate Adjusted Sample Skewness
Use Collabora SKEW with at least three numeric observations and distinguish its adjustment from population skewness.
Verification status: See the evidence and test scope described in this guide. How formula examples are checked
Quick answer: for 2, 3, 1, 6, 8 and 5 in A2:A7, =SKEW(A2:A7) has independently expected result approximately 0.319046413.
Syntax and the adjusted statistic
=SKEW(Argument1 [; Argument2 ...])
The dedicated Collabora help permits up to 255 arguments, including references and ranges, and requires at least three numeric values. Text and empty cells within a referenced range are ignored.
The current source standardizes centered values using the sample standard deviation, cubes those standardized values and applies the factor n/((n-1)*(n-2)). This is adjusted sample skewness, distinct from the population-moment calculation supplied by SKEWP.
A positive result indicates positive third-moment asymmetry under this definition; a negative result indicates the opposite direction. A zero result is compatible with symmetry but does not prove that an arbitrary distribution is symmetric.
Example 1: reproduce an ordinary numeric range
Use the six values from the quick answer. Their sum is 25, mean 25/6, approximately 4.166666667, and sample variance 209/30, approximately 6.966666667.
Enter =SKEW(A2:A7) in C2. The independent target is approximately 0.319046413. Keep all six original cells so that the result can be checked against their count and actual stored values.
For a manual check, subtract the mean from each observation, divide each difference by the sample standard deviation, cube the standardized values, sum them and multiply by 6/(5*4). Use unrounded intermediate values rather than a rounded mean copied from a report.
Example 2: compare the population convention
=SKEWP(A2:A7) has independently expected result approximately 0.232998556 for the same six values. The difference is the adjustment convention, not a different dataset.
For n>2 with nonzero spread, the mathematical relationship is SKEW = SKEWPSQRT(n(n-1))/(n-2). At n=6, the factor is approximately 1.369306394, producing the stated adjusted result.
Use the function appropriate to the statistical definition required by the analysis. A population-style skewness is not obtained by simply rounding the adjusted result to fewer decimals.
Advanced application: use symmetry and sign controls
For 1, 2, 3, 4 and 5, the independent SKEW result is 0. The centered positive and negative cubes cancel.
For 1, 1, 1, 1 and 6, the adjusted result is approximately 2.236067977. Negating every value gives the opposite result, approximately -2.236067977, because odd powers change sign.
This sign reversal is a useful audit check. Multiplying by a negative scale changes the direction of asymmetry rather than leaving the statistic unchanged as a fourth-power kurtosis measure would.
Advanced application: distinguish scaling from distribution shape
Adding ten to each observation in the six-value dataset leaves the independent skewness approximately 0.319046413. Multiplying all observations by a positive factor such as two also leaves it unchanged.
Skewness is dimensionless because the standardized deviations remove the measurement unit. A change from meters to centimeters should not itself change this shape statistic, although extreme numeric magnitudes can introduce finite-precision effects.
Keep the original observations and unit conversion visible. An apparent change in skewness after a simple positive unit conversion can reveal inconsistent scaling, missing rows or unintended rounding.
Errors and troubleshooting
The current source rejects fewer than three usable values through a division-by-zero path. If the standard deviation is zero, it follows an illegal-argument path. Exact native localized displays are not certified; a constant range does not have a defined standardized skewness here.
Ordinary text and blanks in a referenced range do not contribute observations. An actual numeric zero does contribute and can change the result substantially. Check COUNT and inspect imported numeric-looking text before assuming the selected cell count equals the sample count.
Skewness does not identify a data-entry error or establish a probability model by itself. Small samples and extreme observations can make the estimate unstable. These examples verify the documented argument mapping and mathematical statistic, not a general statistical conclusion about real data.
Availability and verification
The exact function is documented in Collabora Office 26.04 desktop help. The cited help snapshot was retrieved on 2026-10-13 and hash-verified October 15, 2026. Desktop documentation does not establish Online or mobile support. First introduction and last functioning release are unknown in this evidence.
No native Collabora Office execution occurred. All numerical results are independent mathematical expectations, with English names and semicolon argument separators. Native error displays, all coercions and extreme numerical behavior are not certified.
Implementation statements are corroborated by Collabora current engine source, retrieved October 15, 2026. This is moving-main source evidence, not a pinned desktop 26.04 binary or execution.